Subcutaneous tunnelling of short-term catheters and the choice of site of catheter insertion are the two factors which influence the risk of catheter-related infection. Catheters inserted into a femoral vein have been associated with a higher risk of infection than those inserted into a subclavian vein. However, a meta-analysis comparing the internal jugular and subclavian sites did not report any difference. Regarding tunnelling, two randomized studies showed a benefit in terms of infection rates for the jugular and femoral sites when the blood sampling was not performed via the catheters. For the subclavian site tunnelling does not seem to bring any advantage compared to conventional insertion.